Novel Sensor Geolocation Technology for Detection and Discrimination of Unexploded Ordnance

Reliable and accurate geolocation is essential for robust detection, discrimination and remediation of unexploded ordnance (UXO) and other munitions and explosives-of-concern (MEC). The MEC characterization and remediation activities using the currently available detection and geolocation technologies often yield unsatisfactory results, and are extremely expensive, due mainly to the inability of the current technology to detect all MEC present at a site, and their failure to discriminate between MEC and non-hazardous items. This is a consequence of insufficiently accurate relative geolocation information of the electromagnetic (EM) sensors, since multiple EM images are combined together to produce 3D images of the buried objects, which are often blurred due to the poor sensor geolocation. As a result, most of the cost (90%) of MEC remediation involves excavating targets that pose no threat. Thus, the goal of the research presented here is to design and implement a high-accuracy device that can address the stringent navigation/geolocation requirements of a man-portable EM system in open and impeded environments, and therefore lower the cost of remediation by improving the geolocation accuracy of MEC that will result in better discrimination, by practically eliminating excavation of non-hazardous objects. The proposed system design is based on the tight 'quadruple' integration of the Global Positioning System (GPS), inertial measurement unit system (IMU), terrestrial RF system – pseudolite (PL), and terrestrial laser scanning (TLS) to support high-accuracy navigation for a non-contact mapping system in GPS-challenged environments. The key novel component of the proposed multi-sensor system is the integration of TLS that can provide very high positioning accuracy in a local frame, and thus can support a GPS/INS/PL-based navigation system in achieving both absolute and relative high positioning accuracy in impeded environments. This paper presents the concept design of the quadruple integration system, the algorithmic approach to sensor integration with a special emphasis on TLS integration with GSP/IMU/PL, and the performance assessment based on simulations and real data, where cm-level relative geolocation accuracy is demonstrated.